net/mptcp/pm_netlink.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 /* Multipath TCP
   3  *
   4  * Copyright (c) 2020, Red Hat, Inc.
   5  */
   6
   7 #define pr_fmt(fmt) "MPTCP: " fmt
   8
   9 #include <linux/inet.h>
  10 #include <linux/kernel.h>
  11 #include <net/tcp.h>
  12 #include <net/netns/generic.h>
  13 #include <net/mptcp.h>
  14 #include <net/genetlink.h>
  15 #include <uapi/linux/mptcp.h>
  16
  17 #include "protocol.h"
  18 #include "mib.h"
  19
  20 /* forward declaration */
  21 static struct genl_family mptcp_genl_family;
  22
  23 static int pm_nl_pernet_id;
  24
  25 struct mptcp_pm_addr_entry {
  26         struct list_head        list;
  27         struct mptcp_addr_info  addr;
  28         struct rcu_head         rcu;
  29         struct socket           *lsk;
  30 };
  31
  32 struct mptcp_pm_add_entry {
  33         struct list_head        list;
  34         struct mptcp_addr_info  addr;
  35         struct timer_list       add_timer;
  36         struct mptcp_sock       *sock;
  37         u8                      retrans_times;
  38 };
  39
  40 #define MAX_ADDR_ID             255
  41 #define BITMAP_SZ DIV_ROUND_UP(MAX_ADDR_ID + 1, BITS_PER_LONG)
  42
  43 struct pm_nl_pernet {
  44         /* protects pernet updates */
  45         spinlock_t              lock;
  46         struct list_head        local_addr_list;
  47         unsigned int            addrs;
  48         unsigned int            add_addr_signal_max;
  49         unsigned int            add_addr_accept_max;
  50         unsigned int            local_addr_max;
  51         unsigned int            subflows_max;
  52         unsigned int            next_id;
  53         unsigned long           id_bitmap[BITMAP_SZ];
  54 };
  55
  56 #define MPTCP_PM_ADDR_MAX       8
  57 #define ADD_ADDR_RETRANS_MAX    3
  58
  59 static void mptcp_pm_nl_add_addr_send_ack(struct mptcp_sock *msk);
  60
  61 static bool addresses_equal(const struct mptcp_addr_info *a,
  62                             struct mptcp_addr_info *b, bool use_port)
  63 {
  64         bool addr_equals = false;
  65
  66         if (a->family == b->family) {
  67                 if (a->family == AF_INET)
  68                         addr_equals = a->addr.s_addr == b->addr.s_addr;
  69 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
  70                 else
  71                         addr_equals = !ipv6_addr_cmp(&a->addr6, &b->addr6);
  72         } else if (a->family == AF_INET) {
  73                 if (ipv6_addr_v4mapped(&b->addr6))
  74                         addr_equals = a->addr.s_addr == b->addr6.s6_addr32[3];
  75         } else if (b->family == AF_INET) {
  76                 if (ipv6_addr_v4mapped(&a->addr6))
  77                         addr_equals = a->addr6.s6_addr32[3] == b->addr.s_addr;
  78 #endif
  79         }
  80
  81         if (!addr_equals)
  82                 return false;
  83         if (!use_port)
  84                 return true;
  85
  86         return a->port == b->port;
  87 }
  88
  89 static bool address_zero(const struct mptcp_addr_info *addr)
  90 {
  91         struct mptcp_addr_info zero;
  92
  93         memset(&zero, 0, sizeof(zero));
  94         zero.family = addr->family;
  95
  96         return addresses_equal(addr, &zero, true);
  97 }
  98
  99 static void local_address(const struct sock_common *skc,
 100                           struct mptcp_addr_info *addr)
 101 {
 102         addr->family = skc->skc_family;
 103         addr->port = htons(skc->skc_num);
 104         if (addr->family == AF_INET)
 105                 addr->addr.s_addr = skc->skc_rcv_saddr;
 106 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 107         else if (addr->family == AF_INET6)
 108                 addr->addr6 = skc->skc_v6_rcv_saddr;
 109 #endif
 110 }
 111
 112 static void remote_address(const struct sock_common *skc,
 113                            struct mptcp_addr_info *addr)
 114 {
 115         addr->family = skc->skc_family;
 116         addr->port = skc->skc_dport;
 117         if (addr->family == AF_INET)
 118                 addr->addr.s_addr = skc->skc_daddr;
 119 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 120         else if (addr->family == AF_INET6)
 121                 addr->addr6 = skc->skc_v6_daddr;
 122 #endif
 123 }
 124
 125 static bool lookup_subflow_by_saddr(const struct list_head *list,
 126                                     struct mptcp_addr_info *saddr)
 127 {
 128         struct mptcp_subflow_context *subflow;
 129         struct mptcp_addr_info cur;
 130         struct sock_common *skc;
 131
 132         list_for_each_entry(subflow, list, node) {
 133                 skc = (struct sock_common *)mptcp_subflow_tcp_sock(subflow);
 134
 135                 local_address(skc, &cur);
 136                 if (addresses_equal(&cur, saddr, saddr->port))
 137                         return true;
 138         }
 139
 140         return false;
 141 }
 142
 143 static struct mptcp_pm_addr_entry *
 144 select_local_address(const struct pm_nl_pernet *pernet,
 145                      struct mptcp_sock *msk)
 146 {
 147         struct mptcp_pm_addr_entry *entry, *ret = NULL;
 148         struct sock *sk = (struct sock *)msk;
 149
 150         msk_owned_by_me(msk);
 151
 152         rcu_read_lock();
 153         __mptcp_flush_join_list(msk);
 154         list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) {
 155                 if (!(entry->addr.flags & MPTCP_PM_ADDR_FLAG_SUBFLOW))
 156                         continue;
 157
 158                 if (entry->addr.family != sk->sk_family) {
 159 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 160                         if ((entry->addr.family == AF_INET &&
 161                              !ipv6_addr_v4mapped(&sk->sk_v6_daddr)) ||
 162                             (sk->sk_family == AF_INET &&
 163                              !ipv6_addr_v4mapped(&entry->addr.addr6)))
 164 #endif
 165                                 continue;
 166                 }
 167
 168                 /* avoid any address already in use by subflows and
 169                  * pending join
 170                  */
 171                 if (!lookup_subflow_by_saddr(&msk->conn_list, &entry->addr)) {
 172                         ret = entry;
 173                         break;
 174                 }
 175         }
 176         rcu_read_unlock();
 177         return ret;
 178 }
 179
 180 static struct mptcp_pm_addr_entry *
 181 select_signal_address(struct pm_nl_pernet *pernet, unsigned int pos)
 182 {
 183         struct mptcp_pm_addr_entry *entry, *ret = NULL;
 184         int i = 0;
 185
 186         rcu_read_lock();
 187         /* do not keep any additional per socket state, just signal
 188          * the address list in order.
 189          * Note: removal from the local address list during the msk life-cycle
 190          * can lead to additional addresses not being announced.
 191          */
 192         list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) {
 193                 if (!(entry->addr.flags & MPTCP_PM_ADDR_FLAG_SIGNAL))
 194                         continue;
 195                 if (i++ == pos) {
 196                         ret = entry;
 197                         break;
 198                 }
 199         }
 200         rcu_read_unlock();
 201         return ret;
 202 }
 203
 204 unsigned int mptcp_pm_get_add_addr_signal_max(struct mptcp_sock *msk)
 205 {
 206         struct pm_nl_pernet *pernet;
 207
 208         pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 209         return READ_ONCE(pernet->add_addr_signal_max);
 210 }
 211 EXPORT_SYMBOL_GPL(mptcp_pm_get_add_addr_signal_max);
 212
 213 unsigned int mptcp_pm_get_add_addr_accept_max(struct mptcp_sock *msk)
 214 {
 215         struct pm_nl_pernet *pernet;
 216
 217         pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 218         return READ_ONCE(pernet->add_addr_accept_max);
 219 }
 220 EXPORT_SYMBOL_GPL(mptcp_pm_get_add_addr_accept_max);
 221
 222 unsigned int mptcp_pm_get_subflows_max(struct mptcp_sock *msk)
 223 {
 224         struct pm_nl_pernet *pernet;
 225
 226         pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 227         return READ_ONCE(pernet->subflows_max);
 228 }
 229 EXPORT_SYMBOL_GPL(mptcp_pm_get_subflows_max);
 230
 231 unsigned int mptcp_pm_get_local_addr_max(struct mptcp_sock *msk)
 232 {
 233         struct pm_nl_pernet *pernet;
 234
 235         pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 236         return READ_ONCE(pernet->local_addr_max);
 237 }
 238 EXPORT_SYMBOL_GPL(mptcp_pm_get_local_addr_max);
 239
 240 static void check_work_pending(struct mptcp_sock *msk)
 241 {
 242         if (msk->pm.add_addr_signaled == mptcp_pm_get_add_addr_signal_max(msk) &&
 243             (msk->pm.local_addr_used == mptcp_pm_get_local_addr_max(msk) ||
 244              msk->pm.subflows == mptcp_pm_get_subflows_max(msk)))
 245                 WRITE_ONCE(msk->pm.work_pending, false);
 246 }
 247
 248 static struct mptcp_pm_add_entry *
 249 lookup_anno_list_by_saddr(struct mptcp_sock *msk,
 250                           struct mptcp_addr_info *addr)
 251 {
 252         struct mptcp_pm_add_entry *entry;
 253
 254         lockdep_assert_held(&msk->pm.lock);
 255
 256         list_for_each_entry(entry, &msk->pm.anno_list, list) {
 257                 if (addresses_equal(&entry->addr, addr, true))
 258                         return entry;
 259         }
 260
 261         return NULL;
 262 }
 263
 264 bool mptcp_pm_sport_in_anno_list(struct mptcp_sock *msk, const struct sock *sk)
 265 {
 266         struct mptcp_pm_add_entry *entry;
 267         struct mptcp_addr_info saddr;
 268         bool ret = false;
 269
 270         local_address((struct sock_common *)sk, &saddr);
 271
 272         spin_lock_bh(&msk->pm.lock);
 273         list_for_each_entry(entry, &msk->pm.anno_list, list) {
 274                 if (addresses_equal(&entry->addr, &saddr, true)) {
 275                         ret = true;
 276                         goto out;
 277                 }
 278         }
 279
 280 out:
 281         spin_unlock_bh(&msk->pm.lock);
 282         return ret;
 283 }
 284
 285 static void mptcp_pm_add_timer(struct timer_list *timer)
 286 {
 287         struct mptcp_pm_add_entry *entry = from_timer(entry, timer, add_timer);
 288         struct mptcp_sock *msk = entry->sock;
 289         struct sock *sk = (struct sock *)msk;
 290
 291         pr_debug("msk=%p", msk);
 292
 293         if (!msk)
 294                 return;
 295
 296         if (inet_sk_state_load(sk) == TCP_CLOSE)
 297                 return;
 298
 299         if (!entry->addr.id)
 300                 return;
 301
 302         if (mptcp_pm_should_add_signal(msk)) {
 303                 sk_reset_timer(sk, timer, jiffies + TCP_RTO_MAX / 8);
 304                 goto out;
 305         }
 306
 307         spin_lock_bh(&msk->pm.lock);
 308
 309         if (!mptcp_pm_should_add_signal(msk)) {
 310                 pr_debug("retransmit ADD_ADDR id=%d", entry->addr.id);
 311                 mptcp_pm_announce_addr(msk, &entry->addr, false, entry->addr.port);
 312                 mptcp_pm_add_addr_send_ack(msk);
 313                 entry->retrans_times++;
 314         }
 315
 316         if (entry->retrans_times < ADD_ADDR_RETRANS_MAX)
 317                 sk_reset_timer(sk, timer,
 318                                jiffies + mptcp_get_add_addr_timeout(sock_net(sk)));
 319
 320         spin_unlock_bh(&msk->pm.lock);
 321
 322 out:
 323         __sock_put(sk);
 324 }
 325
 326 struct mptcp_pm_add_entry *
 327 mptcp_pm_del_add_timer(struct mptcp_sock *msk,
 328                        struct mptcp_addr_info *addr)
 329 {
 330         struct mptcp_pm_add_entry *entry;
 331         struct sock *sk = (struct sock *)msk;
 332
 333         spin_lock_bh(&msk->pm.lock);
 334         entry = lookup_anno_list_by_saddr(msk, addr);
 335         if (entry)
 336                 entry->retrans_times = ADD_ADDR_RETRANS_MAX;
 337         spin_unlock_bh(&msk->pm.lock);
 338
 339         if (entry)
 340                 sk_stop_timer_sync(sk, &entry->add_timer);
 341
 342         return entry;
 343 }
 344
 345 static bool mptcp_pm_alloc_anno_list(struct mptcp_sock *msk,
 346                                      struct mptcp_pm_addr_entry *entry)
 347 {
 348         struct mptcp_pm_add_entry *add_entry = NULL;
 349         struct sock *sk = (struct sock *)msk;
 350         struct net *net = sock_net(sk);
 351
 352         lockdep_assert_held(&msk->pm.lock);
 353
 354         if (lookup_anno_list_by_saddr(msk, &entry->addr))
 355                 return false;
 356
 357         add_entry = kmalloc(sizeof(*add_entry), GFP_ATOMIC);
 358         if (!add_entry)
 359                 return false;
 360
 361         list_add(&add_entry->list, &msk->pm.anno_list);
 362
 363         add_entry->addr = entry->addr;
 364         add_entry->sock = msk;
 365         add_entry->retrans_times = 0;
 366
 367         timer_setup(&add_entry->add_timer, mptcp_pm_add_timer, 0);
 368         sk_reset_timer(sk, &add_entry->add_timer,
 369                        jiffies + mptcp_get_add_addr_timeout(net));
 370
 371         return true;
 372 }
 373
 374 void mptcp_pm_free_anno_list(struct mptcp_sock *msk)
 375 {
 376         struct mptcp_pm_add_entry *entry, *tmp;
 377         struct sock *sk = (struct sock *)msk;
 378         LIST_HEAD(free_list);
 379
 380         pr_debug("msk=%p", msk);
 381
 382         spin_lock_bh(&msk->pm.lock);
 383         list_splice_init(&msk->pm.anno_list, &free_list);
 384         spin_unlock_bh(&msk->pm.lock);
 385
 386         list_for_each_entry_safe(entry, tmp, &free_list, list) {
 387                 sk_stop_timer_sync(sk, &entry->add_timer);
 388                 kfree(entry);
 389         }
 390 }
 391
 392 static void mptcp_pm_create_subflow_or_signal_addr(struct mptcp_sock *msk)
 393 {
 394         struct sock *sk = (struct sock *)msk;
 395         struct mptcp_pm_addr_entry *local;
 396         unsigned int add_addr_signal_max;
 397         unsigned int local_addr_max;
 398         struct pm_nl_pernet *pernet;
 399         unsigned int subflows_max;
 400
 401         pernet = net_generic(sock_net(sk), pm_nl_pernet_id);
 402
 403         add_addr_signal_max = mptcp_pm_get_add_addr_signal_max(msk);
 404         local_addr_max = mptcp_pm_get_local_addr_max(msk);
 405         subflows_max = mptcp_pm_get_subflows_max(msk);
 406
 407         pr_debug("local %d:%d signal %d:%d subflows %d:%d\n",
 408                  msk->pm.local_addr_used, local_addr_max,
 409                  msk->pm.add_addr_signaled, add_addr_signal_max,
 410                  msk->pm.subflows, subflows_max);
 411
 412         /* check first for announce */
 413         if (msk->pm.add_addr_signaled < add_addr_signal_max) {
 414                 local = select_signal_address(pernet,
 415                                               msk->pm.add_addr_signaled);
 416
 417                 if (local) {
 418                         if (mptcp_pm_alloc_anno_list(msk, local)) {
 419                                 msk->pm.add_addr_signaled++;
 420                                 mptcp_pm_announce_addr(msk, &local->addr, false, local->addr.port);
 421                                 mptcp_pm_nl_add_addr_send_ack(msk);
 422                         }
 423                 } else {
 424                         /* pick failed, avoid fourther attempts later */
 425                         msk->pm.local_addr_used = add_addr_signal_max;
 426                 }
 427
 428                 check_work_pending(msk);
 429         }
 430
 431         /* check if should create a new subflow */
 432         if (msk->pm.local_addr_used < local_addr_max &&
 433             msk->pm.subflows < subflows_max) {
 434                 local = select_local_address(pernet, msk);
 435                 if (local) {
 436                         struct mptcp_addr_info remote = { 0 };
 437
 438                         msk->pm.local_addr_used++;
 439                         msk->pm.subflows++;
 440                         check_work_pending(msk);
 441                         remote_address((struct sock_common *)sk, &remote);
 442                         spin_unlock_bh(&msk->pm.lock);
 443                         __mptcp_subflow_connect(sk, &local->addr, &remote);
 444                         spin_lock_bh(&msk->pm.lock);
 445                         return;
 446                 }
 447
 448                 /* lookup failed, avoid fourther attempts later */
 449                 msk->pm.local_addr_used = local_addr_max;
 450                 check_work_pending(msk);
 451         }
 452 }
 453
 454 static void mptcp_pm_nl_fully_established(struct mptcp_sock *msk)
 455 {
 456         mptcp_pm_create_subflow_or_signal_addr(msk);
 457 }
 458
 459 static void mptcp_pm_nl_subflow_established(struct mptcp_sock *msk)
 460 {
 461         mptcp_pm_create_subflow_or_signal_addr(msk);
 462 }
 463
 464 static void mptcp_pm_nl_add_addr_received(struct mptcp_sock *msk)
 465 {
 466         struct sock *sk = (struct sock *)msk;
 467         unsigned int add_addr_accept_max;
 468         struct mptcp_addr_info remote;
 469         struct mptcp_addr_info local;
 470         unsigned int subflows_max;
 471         bool use_port = false;
 472
 473         add_addr_accept_max = mptcp_pm_get_add_addr_accept_max(msk);
 474         subflows_max = mptcp_pm_get_subflows_max(msk);
 475
 476         pr_debug("accepted %d:%d remote family %d",
 477                  msk->pm.add_addr_accepted, add_addr_accept_max,
 478                  msk->pm.remote.family);
 479         msk->pm.add_addr_accepted++;
 480         msk->pm.subflows++;
 481         if (msk->pm.add_addr_accepted >= add_addr_accept_max ||
 482             msk->pm.subflows >= subflows_max)
 483                 WRITE_ONCE(msk->pm.accept_addr, false);
 484
 485         /* connect to the specified remote address, using whatever
 486          * local address the routing configuration will pick.
 487          */
 488         remote = msk->pm.remote;
 489         if (!remote.port)
 490                 remote.port = sk->sk_dport;
 491         else
 492                 use_port = true;
 493         memset(&local, 0, sizeof(local));
 494         local.family = remote.family;
 495
 496         spin_unlock_bh(&msk->pm.lock);
 497         __mptcp_subflow_connect(sk, &local, &remote);
 498         spin_lock_bh(&msk->pm.lock);
 499
 500         mptcp_pm_announce_addr(msk, &remote, true, use_port);
 501         mptcp_pm_nl_add_addr_send_ack(msk);
 502 }
 503
 504 static void mptcp_pm_nl_add_addr_send_ack(struct mptcp_sock *msk)
 505 {
 506         struct mptcp_subflow_context *subflow;
 507
 508         msk_owned_by_me(msk);
 509         lockdep_assert_held(&msk->pm.lock);
 510
 511         if (!mptcp_pm_should_add_signal(msk))
 512                 return;
 513
 514         __mptcp_flush_join_list(msk);
 515         subflow = list_first_entry_or_null(&msk->conn_list, typeof(*subflow), node);
 516         if (subflow) {
 517                 struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 518                 u8 add_addr;
 519
 520                 spin_unlock_bh(&msk->pm.lock);
 521                 pr_debug("send ack for add_addr%s%s",
 522                          mptcp_pm_should_add_signal_ipv6(msk) ? " [ipv6]" : "",
 523                          mptcp_pm_should_add_signal_port(msk) ? " [port]" : "");
 524
 525                 lock_sock(ssk);
 526                 tcp_send_ack(ssk);
 527                 release_sock(ssk);
 528                 spin_lock_bh(&msk->pm.lock);
 529
 530                 add_addr = READ_ONCE(msk->pm.addr_signal);
 531                 if (mptcp_pm_should_add_signal_ipv6(msk))
 532                         add_addr &= ~BIT(MPTCP_ADD_ADDR_IPV6);
 533                 if (mptcp_pm_should_add_signal_port(msk))
 534                         add_addr &= ~BIT(MPTCP_ADD_ADDR_PORT);
 535                 WRITE_ONCE(msk->pm.addr_signal, add_addr);
 536         }
 537 }
 538
 539 int mptcp_pm_nl_mp_prio_send_ack(struct mptcp_sock *msk,
 540                                  struct mptcp_addr_info *addr,
 541                                  u8 bkup)
 542 {
 543         struct mptcp_subflow_context *subflow;
 544
 545         pr_debug("bkup=%d", bkup);
 546
 547         mptcp_for_each_subflow(msk, subflow) {
 548                 struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 549                 struct sock *sk = (struct sock *)msk;
 550                 struct mptcp_addr_info local;
 551
 552                 local_address((struct sock_common *)ssk, &local);
 553                 if (!addresses_equal(&local, addr, addr->port))
 554                         continue;
 555
 556                 subflow->backup = bkup;
 557                 subflow->send_mp_prio = 1;
 558                 subflow->request_bkup = bkup;
 559                 __MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_MPPRIOTX);
 560
 561                 spin_unlock_bh(&msk->pm.lock);
 562                 pr_debug("send ack for mp_prio");
 563                 lock_sock(ssk);
 564                 tcp_send_ack(ssk);
 565                 release_sock(ssk);
 566                 spin_lock_bh(&msk->pm.lock);
 567
 568                 return 0;
 569         }
 570
 571         return -EINVAL;
 572 }
 573
 574 static void mptcp_pm_nl_rm_addr_received(struct mptcp_sock *msk)
 575 {
 576         struct mptcp_subflow_context *subflow, *tmp;
 577         struct sock *sk = (struct sock *)msk;
 578         u8 i;
 579
 580         pr_debug("address rm_list_nr %d", msk->pm.rm_list_rx.nr);
 581
 582         msk_owned_by_me(msk);
 583
 584         if (!msk->pm.rm_list_rx.nr)
 585                 return;
 586
 587         if (list_empty(&msk->conn_list))
 588                 return;
 589
 590         for (i = 0; i < msk->pm.rm_list_rx.nr; i++) {
 591                 list_for_each_entry_safe(subflow, tmp, &msk->conn_list, node) {
 592                         struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 593                         int how = RCV_SHUTDOWN | SEND_SHUTDOWN;
 594
 595                         if (msk->pm.rm_list_rx.ids[i] != subflow->remote_id)
 596                                 continue;
 597
 598                         pr_debug(" -> address rm_list_ids[%d]=%u", i, msk->pm.rm_list_rx.ids[i]);
 599                         spin_unlock_bh(&msk->pm.lock);
 600                         mptcp_subflow_shutdown(sk, ssk, how);
 601                         mptcp_close_ssk(sk, ssk, subflow);
 602                         spin_lock_bh(&msk->pm.lock);
 603
 604                         msk->pm.add_addr_accepted--;
 605                         msk->pm.subflows--;
 606                         WRITE_ONCE(msk->pm.accept_addr, true);
 607
 608                         __MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_RMADDR);
 609
 610                         break;
 611                 }
 612         }
 613 }
 614
 615 void mptcp_pm_nl_work(struct mptcp_sock *msk)
 616 {
 617         struct mptcp_pm_data *pm = &msk->pm;
 618
 619         msk_owned_by_me(msk);
 620
 621         spin_lock_bh(&msk->pm.lock);
 622
 623         pr_debug("msk=%p status=%x", msk, pm->status);
 624         if (pm->status & BIT(MPTCP_PM_ADD_ADDR_RECEIVED)) {
 625                 pm->status &= ~BIT(MPTCP_PM_ADD_ADDR_RECEIVED);
 626                 mptcp_pm_nl_add_addr_received(msk);
 627         }
 628         if (pm->status & BIT(MPTCP_PM_ADD_ADDR_SEND_ACK)) {
 629                 pm->status &= ~BIT(MPTCP_PM_ADD_ADDR_SEND_ACK);
 630                 mptcp_pm_nl_add_addr_send_ack(msk);
 631         }
 632         if (pm->status & BIT(MPTCP_PM_RM_ADDR_RECEIVED)) {
 633                 pm->status &= ~BIT(MPTCP_PM_RM_ADDR_RECEIVED);
 634                 mptcp_pm_nl_rm_addr_received(msk);
 635         }
 636         if (pm->status & BIT(MPTCP_PM_ESTABLISHED)) {
 637                 pm->status &= ~BIT(MPTCP_PM_ESTABLISHED);
 638                 mptcp_pm_nl_fully_established(msk);
 639         }
 640         if (pm->status & BIT(MPTCP_PM_SUBFLOW_ESTABLISHED)) {
 641                 pm->status &= ~BIT(MPTCP_PM_SUBFLOW_ESTABLISHED);
 642                 mptcp_pm_nl_subflow_established(msk);
 643         }
 644
 645         spin_unlock_bh(&msk->pm.lock);
 646 }
 647
 648 void mptcp_pm_nl_rm_subflow_received(struct mptcp_sock *msk,
 649                                      const struct mptcp_rm_list *rm_list)
 650 {
 651         struct mptcp_subflow_context *subflow, *tmp;
 652         struct sock *sk = (struct sock *)msk;
 653         u8 i;
 654
 655         pr_debug("subflow rm_list_nr %d", rm_list->nr);
 656
 657         msk_owned_by_me(msk);
 658
 659         if (!rm_list->nr)
 660                 return;
 661
 662         if (list_empty(&msk->conn_list))
 663                 return;
 664
 665         for (i = 0; i < rm_list->nr; i++) {
 666                 list_for_each_entry_safe(subflow, tmp, &msk->conn_list, node) {
 667                         struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 668                         int how = RCV_SHUTDOWN | SEND_SHUTDOWN;
 669
 670                         if (rm_list->ids[i] != subflow->local_id)
 671                                 continue;
 672
 673                         pr_debug(" -> subflow rm_list_ids[%d]=%u", i, rm_list->ids[i]);
 674                         spin_unlock_bh(&msk->pm.lock);
 675                         mptcp_subflow_shutdown(sk, ssk, how);
 676                         mptcp_close_ssk(sk, ssk, subflow);
 677                         spin_lock_bh(&msk->pm.lock);
 678
 679                         msk->pm.local_addr_used--;
 680                         msk->pm.subflows--;
 681
 682                         __MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_RMSUBFLOW);
 683
 684                         break;
 685                 }
 686         }
 687 }
 688
 689 static bool address_use_port(struct mptcp_pm_addr_entry *entry)
 690 {
 691         return (entry->addr.flags &
 692                 (MPTCP_PM_ADDR_FLAG_SIGNAL | MPTCP_PM_ADDR_FLAG_SUBFLOW)) ==
 693                 MPTCP_PM_ADDR_FLAG_SIGNAL;
 694 }
 695
 696 static int mptcp_pm_nl_append_new_local_addr(struct pm_nl_pernet *pernet,
 697                                              struct mptcp_pm_addr_entry *entry)
 698 {
 699         struct mptcp_pm_addr_entry *cur;
 700         unsigned int addr_max;
 701         int ret = -EINVAL;
 702
 703         spin_lock_bh(&pernet->lock);
 704         /* to keep the code simple, don't do IDR-like allocation for address ID,
 705          * just bail when we exceed limits
 706          */
 707         if (pernet->next_id == MAX_ADDR_ID)
 708                 pernet->next_id = 1;
 709         if (pernet->addrs >= MPTCP_PM_ADDR_MAX)
 710                 goto out;
 711         if (test_bit(entry->addr.id, pernet->id_bitmap))
 712                 goto out;
 713
 714         /* do not insert duplicate address, differentiate on port only
 715          * singled addresses
 716          */
 717         list_for_each_entry(cur, &pernet->local_addr_list, list) {
 718                 if (addresses_equal(&cur->addr, &entry->addr,
 719                                     address_use_port(entry) &&
 720                                     address_use_port(cur)))
 721                         goto out;
 722         }
 723
 724         if (!entry->addr.id) {
 725 find_next:
 726                 entry->addr.id = find_next_zero_bit(pernet->id_bitmap,
 727                                                     MAX_ADDR_ID + 1,
 728                                                     pernet->next_id);
 729                 if ((!entry->addr.id || entry->addr.id > MAX_ADDR_ID) &&
 730                     pernet->next_id != 1) {
 731                         pernet->next_id = 1;
 732                         goto find_next;
 733                 }
 734         }
 735
 736         if (!entry->addr.id || entry->addr.id > MAX_ADDR_ID)
 737                 goto out;
 738
 739         __set_bit(entry->addr.id, pernet->id_bitmap);
 740         if (entry->addr.id > pernet->next_id)
 741                 pernet->next_id = entry->addr.id;
 742
 743         if (entry->addr.flags & MPTCP_PM_ADDR_FLAG_SIGNAL) {
 744                 addr_max = pernet->add_addr_signal_max;
 745                 WRITE_ONCE(pernet->add_addr_signal_max, addr_max + 1);
 746         }
 747         if (entry->addr.flags & MPTCP_PM_ADDR_FLAG_SUBFLOW) {
 748                 addr_max = pernet->local_addr_max;
 749                 WRITE_ONCE(pernet->local_addr_max, addr_max + 1);
 750         }
 751
 752         pernet->addrs++;
 753         list_add_tail_rcu(&entry->list, &pernet->local_addr_list);
 754         ret = entry->addr.id;
 755
 756 out:
 757         spin_unlock_bh(&pernet->lock);
 758         return ret;
 759 }
 760
 761 static int mptcp_pm_nl_create_listen_socket(struct sock *sk,
 762                                             struct mptcp_pm_addr_entry *entry)
 763 {
 764         struct sockaddr_storage addr;
 765         struct mptcp_sock *msk;
 766         struct socket *ssock;
 767         int backlog = 1024;
 768         int err;
 769
 770         err = sock_create_kern(sock_net(sk), entry->addr.family,
 771                                SOCK_STREAM, IPPROTO_MPTCP, &entry->lsk);
 772         if (err)
 773                 return err;
 774
 775         msk = mptcp_sk(entry->lsk->sk);
 776         if (!msk) {
 777                 err = -EINVAL;
 778                 goto out;
 779         }
 780
 781         ssock = __mptcp_nmpc_socket(msk);
 782         if (!ssock) {
 783                 err = -EINVAL;
 784                 goto out;
 785         }
 786
 787         mptcp_info2sockaddr(&entry->addr, &addr, entry->addr.family);
 788         err = kernel_bind(ssock, (struct sockaddr *)&addr,
 789                           sizeof(struct sockaddr_in));
 790         if (err) {
 791                 pr_warn("kernel_bind error, err=%d", err);
 792                 goto out;
 793         }
 794
 795         err = kernel_listen(ssock, backlog);
 796         if (err) {
 797                 pr_warn("kernel_listen error, err=%d", err);
 798                 goto out;
 799         }
 800
 801         return 0;
 802
 803 out:
 804         sock_release(entry->lsk);
 805         return err;
 806 }
 807
 808 int mptcp_pm_nl_get_local_id(struct mptcp_sock *msk, struct sock_common *skc)
 809 {
 810         struct mptcp_pm_addr_entry *entry;
 811         struct mptcp_addr_info skc_local;
 812         struct mptcp_addr_info msk_local;
 813         struct pm_nl_pernet *pernet;
 814         int ret = -1;
 815
 816         if (WARN_ON_ONCE(!msk))
 817                 return -1;
 818
 819         /* The 0 ID mapping is defined by the first subflow, copied into the msk
 820          * addr
 821          */
 822         local_address((struct sock_common *)msk, &msk_local);
 823         local_address((struct sock_common *)skc, &skc_local);
 824         if (addresses_equal(&msk_local, &skc_local, false))
 825                 return 0;
 826
 827         if (address_zero(&skc_local))
 828                 return 0;
 829
 830         pernet = net_generic(sock_net((struct sock *)msk), pm_nl_pernet_id);
 831
 832         rcu_read_lock();
 833         list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) {
 834                 if (addresses_equal(&entry->addr, &skc_local, entry->addr.port)) {
 835                         ret = entry->addr.id;
 836                         break;
 837                 }
 838         }
 839         rcu_read_unlock();
 840         if (ret >= 0)
 841                 return ret;
 842
 843         /* address not found, add to local list */
 844         entry = kmalloc(sizeof(*entry), GFP_ATOMIC);
 845         if (!entry)
 846                 return -ENOMEM;
 847
 848         entry->addr = skc_local;
 849         entry->addr.ifindex = 0;
 850         entry->addr.flags = 0;
 851         entry->addr.id = 0;
 852         entry->addr.port = 0;
 853         entry->lsk = NULL;
 854         ret = mptcp_pm_nl_append_new_local_addr(pernet, entry);
 855         if (ret < 0)
 856                 kfree(entry);
 857
 858         return ret;
 859 }
 860
 861 void mptcp_pm_nl_data_init(struct mptcp_sock *msk)
 862 {
 863         struct mptcp_pm_data *pm = &msk->pm;
 864         bool subflows;
 865
 866         subflows = !!mptcp_pm_get_subflows_max(msk);
 867         WRITE_ONCE(pm->work_pending, (!!mptcp_pm_get_local_addr_max(msk) && subflows) ||
 868                    !!mptcp_pm_get_add_addr_signal_max(msk));
 869         WRITE_ONCE(pm->accept_addr, !!mptcp_pm_get_add_addr_accept_max(msk) && subflows);
 870         WRITE_ONCE(pm->accept_subflow, subflows);
 871 }
 872
 873 #define MPTCP_PM_CMD_GRP_OFFSET       0
 874 #define MPTCP_PM_EV_GRP_OFFSET        1
 875
 876 static const struct genl_multicast_group mptcp_pm_mcgrps[] = {
 877         [MPTCP_PM_CMD_GRP_OFFSET]       = { .name = MPTCP_PM_CMD_GRP_NAME, },
 878         [MPTCP_PM_EV_GRP_OFFSET]        = { .name = MPTCP_PM_EV_GRP_NAME,
 879                                             .flags = GENL_UNS_ADMIN_PERM,
 880                                           },
 881 };
 882
 883 static const struct nla_policy
 884 mptcp_pm_addr_policy[MPTCP_PM_ADDR_ATTR_MAX + 1] = {
 885         [MPTCP_PM_ADDR_ATTR_FAMILY]     = { .type       = NLA_U16,      },
 886         [MPTCP_PM_ADDR_ATTR_ID]         = { .type       = NLA_U8,       },
 887         [MPTCP_PM_ADDR_ATTR_ADDR4]      = { .type       = NLA_U32,      },
 888         [MPTCP_PM_ADDR_ATTR_ADDR6]      =
 889                 NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)),
 890         [MPTCP_PM_ADDR_ATTR_PORT]       = { .type       = NLA_U16       },
 891         [MPTCP_PM_ADDR_ATTR_FLAGS]      = { .type       = NLA_U32       },
 892         [MPTCP_PM_ADDR_ATTR_IF_IDX]     = { .type       = NLA_S32       },
 893 };
 894
 895 static const struct nla_policy mptcp_pm_policy[MPTCP_PM_ATTR_MAX + 1] = {
 896         [MPTCP_PM_ATTR_ADDR]            =
 897                                         NLA_POLICY_NESTED(mptcp_pm_addr_policy),
 898         [MPTCP_PM_ATTR_RCV_ADD_ADDRS]   = { .type       = NLA_U32,      },
 899         [MPTCP_PM_ATTR_SUBFLOWS]        = { .type       = NLA_U32,      },
 900 };
 901
 902 static int mptcp_pm_family_to_addr(int family)
 903 {
 904 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 905         if (family == AF_INET6)
 906                 return MPTCP_PM_ADDR_ATTR_ADDR6;
 907 #endif
 908         return MPTCP_PM_ADDR_ATTR_ADDR4;
 909 }
 910
 911 static int mptcp_pm_parse_addr(struct nlattr *attr, struct genl_info *info,
 912                                bool require_family,
 913                                struct mptcp_pm_addr_entry *entry)
 914 {
 915         struct nlattr *tb[MPTCP_PM_ADDR_ATTR_MAX + 1];
 916         int err, addr_addr;
 917
 918         if (!attr) {
 919                 GENL_SET_ERR_MSG(info, "missing address info");
 920                 return -EINVAL;
 921         }
 922
 923         /* no validation needed - was already done via nested policy */
 924         err = nla_parse_nested_deprecated(tb, MPTCP_PM_ADDR_ATTR_MAX, attr,
 925                                           mptcp_pm_addr_policy, info->extack);
 926         if (err)
 927                 return err;
 928
 929         memset(entry, 0, sizeof(*entry));
 930         if (!tb[MPTCP_PM_ADDR_ATTR_FAMILY]) {
 931                 if (!require_family)
 932                         goto skip_family;
 933
 934                 NL_SET_ERR_MSG_ATTR(info->extack, attr,
 935                                     "missing family");
 936                 return -EINVAL;
 937         }
 938
 939         entry->addr.family = nla_get_u16(tb[MPTCP_PM_ADDR_ATTR_FAMILY]);
 940         if (entry->addr.family != AF_INET
 941 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 942             && entry->addr.family != AF_INET6
 943 #endif
 944             ) {
 945                 NL_SET_ERR_MSG_ATTR(info->extack, attr,
 946                                     "unknown address family");
 947                 return -EINVAL;
 948         }
 949         addr_addr = mptcp_pm_family_to_addr(entry->addr.family);
 950         if (!tb[addr_addr]) {
 951                 NL_SET_ERR_MSG_ATTR(info->extack, attr,
 952                                     "missing address data");
 953                 return -EINVAL;
 954         }
 955
 956 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
 957         if (entry->addr.family == AF_INET6)
 958                 entry->addr.addr6 = nla_get_in6_addr(tb[addr_addr]);
 959         else
 960 #endif
 961                 entry->addr.addr.s_addr = nla_get_in_addr(tb[addr_addr]);
 962
 963 skip_family:
 964         if (tb[MPTCP_PM_ADDR_ATTR_IF_IDX]) {
 965                 u32 val = nla_get_s32(tb[MPTCP_PM_ADDR_ATTR_IF_IDX]);
 966
 967                 entry->addr.ifindex = val;
 968         }
 969
 970         if (tb[MPTCP_PM_ADDR_ATTR_ID])
 971                 entry->addr.id = nla_get_u8(tb[MPTCP_PM_ADDR_ATTR_ID]);
 972
 973         if (tb[MPTCP_PM_ADDR_ATTR_FLAGS])
 974                 entry->addr.flags = nla_get_u32(tb[MPTCP_PM_ADDR_ATTR_FLAGS]);
 975
 976         if (tb[MPTCP_PM_ADDR_ATTR_PORT])
 977                 entry->addr.port = htons(nla_get_u16(tb[MPTCP_PM_ADDR_ATTR_PORT]));
 978
 979         return 0;
 980 }
 981
 982 static struct pm_nl_pernet *genl_info_pm_nl(struct genl_info *info)
 983 {
 984         return net_generic(genl_info_net(info), pm_nl_pernet_id);
 985 }
 986
 987 static int mptcp_nl_add_subflow_or_signal_addr(struct net *net)
 988 {
 989         struct mptcp_sock *msk;
 990         long s_slot = 0, s_num = 0;
 991
 992         while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) {
 993                 struct sock *sk = (struct sock *)msk;
 994
 995                 if (!READ_ONCE(msk->fully_established))
 996                         goto next;
 997
 998                 lock_sock(sk);
 999                 spin_lock_bh(&msk->pm.lock);
1000                 mptcp_pm_create_subflow_or_signal_addr(msk);
1001                 spin_unlock_bh(&msk->pm.lock);
1002                 release_sock(sk);
1003
1004 next:
1005                 sock_put(sk);
1006                 cond_resched();
1007         }
1008
1009         return 0;
1010 }
1011
1012 static int mptcp_nl_cmd_add_addr(struct sk_buff *skb, struct genl_info *info)
1013 {
1014         struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR];
1015         struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
1016         struct mptcp_pm_addr_entry addr, *entry;
1017         int ret;
1018
1019         ret = mptcp_pm_parse_addr(attr, info, true, &addr);
1020         if (ret < 0)
1021                 return ret;
1022
1023         entry = kmalloc(sizeof(*entry), GFP_KERNEL);
1024         if (!entry) {
1025                 GENL_SET_ERR_MSG(info, "can't allocate addr");
1026                 return -ENOMEM;
1027         }
1028
1029         *entry = addr;
1030         if (entry->addr.port) {
1031                 ret = mptcp_pm_nl_create_listen_socket(skb->sk, entry);
1032                 if (ret) {
1033                         GENL_SET_ERR_MSG(info, "create listen socket error");
1034                         kfree(entry);
1035                         return ret;
1036                 }
1037         }
1038         ret = mptcp_pm_nl_append_new_local_addr(pernet, entry);
1039         if (ret < 0) {
1040                 GENL_SET_ERR_MSG(info, "too many addresses or duplicate one");
1041                 if (entry->lsk)
1042                         sock_release(entry->lsk);
1043                 kfree(entry);
1044                 return ret;
1045         }
1046
1047         mptcp_nl_add_subflow_or_signal_addr(sock_net(skb->sk));
1048
1049         return 0;
1050 }
1051
1052 static struct mptcp_pm_addr_entry *
1053 __lookup_addr_by_id(struct pm_nl_pernet *pernet, unsigned int id)
1054 {
1055         struct mptcp_pm_addr_entry *entry;
1056
1057         list_for_each_entry(entry, &pernet->local_addr_list, list) {
1058                 if (entry->addr.id == id)
1059                         return entry;
1060         }
1061         return NULL;
1062 }
1063
1064 static bool remove_anno_list_by_saddr(struct mptcp_sock *msk,
1065                                       struct mptcp_addr_info *addr)
1066 {
1067         struct mptcp_pm_add_entry *entry;
1068
1069         entry = mptcp_pm_del_add_timer(msk, addr);
1070         if (entry) {
1071                 list_del(&entry->list);
1072                 kfree(entry);
1073                 return true;
1074         }
1075
1076         return false;
1077 }
1078
1079 static bool mptcp_pm_remove_anno_addr(struct mptcp_sock *msk,
1080                                       struct mptcp_addr_info *addr,
1081                                       bool force)
1082 {
1083         struct mptcp_rm_list list = { .nr = 0 };
1084         bool ret;
1085
1086         list.ids[list.nr++] = addr->id;
1087
1088         ret = remove_anno_list_by_saddr(msk, addr);
1089         if (ret || force) {
1090                 spin_lock_bh(&msk->pm.lock);
1091                 mptcp_pm_remove_addr(msk, &list);
1092                 spin_unlock_bh(&msk->pm.lock);
1093         }
1094         return ret;
1095 }
1096
1097 static int mptcp_nl_remove_subflow_and_signal_addr(struct net *net,
1098                                                    struct mptcp_addr_info *addr)
1099 {
1100         struct mptcp_sock *msk;
1101         long s_slot = 0, s_num = 0;
1102         struct mptcp_rm_list list = { .nr = 0 };
1103
1104         pr_debug("remove_id=%d", addr->id);
1105
1106         list.ids[list.nr++] = addr->id;
1107
1108         while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) {
1109                 struct sock *sk = (struct sock *)msk;
1110                 bool remove_subflow;
1111
1112                 if (list_empty(&msk->conn_list)) {
1113                         mptcp_pm_remove_anno_addr(msk, addr, false);
1114                         goto next;
1115                 }
1116
1117                 lock_sock(sk);
1118                 remove_subflow = lookup_subflow_by_saddr(&msk->conn_list, addr);
1119                 mptcp_pm_remove_anno_addr(msk, addr, remove_subflow);
1120                 if (remove_subflow)
1121                         mptcp_pm_remove_subflow(msk, &list);
1122                 release_sock(sk);
1123
1124 next:
1125                 sock_put(sk);
1126                 cond_resched();
1127         }
1128
1129         return 0;
1130 }
1131
1132 struct addr_entry_release_work {
1133         struct rcu_work rwork;
1134         struct mptcp_pm_addr_entry *entry;
1135 };
1136
1137 static void mptcp_pm_release_addr_entry(struct work_struct *work)
1138 {
1139         struct addr_entry_release_work *w;
1140         struct mptcp_pm_addr_entry *entry;
1141
1142         w = container_of(to_rcu_work(work), struct addr_entry_release_work, rwork);
1143         entry = w->entry;
1144         if (entry) {
1145                 if (entry->lsk)
1146                         sock_release(entry->lsk);
1147                 kfree(entry);
1148         }
1149         kfree(w);
1150 }
1151
1152 static void mptcp_pm_free_addr_entry(struct mptcp_pm_addr_entry *entry)
1153 {
1154         struct addr_entry_release_work *w;
1155
1156         w = kmalloc(sizeof(*w), GFP_ATOMIC);
1157         if (w) {
1158                 INIT_RCU_WORK(&w->rwork, mptcp_pm_release_addr_entry);
1159                 w->entry = entry;
1160                 queue_rcu_work(system_wq, &w->rwork);
1161         }
1162 }
1163
1164 static int mptcp_nl_cmd_del_addr(struct sk_buff *skb, struct genl_info *info)
1165 {
1166         struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR];
1167         struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
1168         struct mptcp_pm_addr_entry addr, *entry;
1169         unsigned int addr_max;
1170         int ret;
1171
1172         ret = mptcp_pm_parse_addr(attr, info, false, &addr);
1173         if (ret < 0)
1174                 return ret;
1175
1176         spin_lock_bh(&pernet->lock);
1177         entry = __lookup_addr_by_id(pernet, addr.addr.id);
1178         if (!entry) {
1179                 GENL_SET_ERR_MSG(info, "address not found");
1180                 spin_unlock_bh(&pernet->lock);
1181                 return -EINVAL;
1182         }
1183         if (entry->addr.flags & MPTCP_PM_ADDR_FLAG_SIGNAL) {
1184                 addr_max = pernet->add_addr_signal_max;
1185                 WRITE_ONCE(pernet->add_addr_signal_max, addr_max - 1);
1186         }
1187         if (entry->addr.flags & MPTCP_PM_ADDR_FLAG_SUBFLOW) {
1188                 addr_max = pernet->local_addr_max;
1189                 WRITE_ONCE(pernet->local_addr_max, addr_max - 1);
1190         }
1191
1192         pernet->addrs--;
1193         list_del_rcu(&entry->list);
1194         __clear_bit(entry->addr.id, pernet->id_bitmap);
1195         spin_unlock_bh(&pernet->lock);
1196
1197         mptcp_nl_remove_subflow_and_signal_addr(sock_net(skb->sk), &entry->addr);
1198         mptcp_pm_free_addr_entry(entry);
1199
1200         return ret;
1201 }
1202
1203 static void mptcp_pm_remove_addrs_and_subflows(struct mptcp_sock *msk,
1204                                                struct list_head *rm_list)
1205 {
1206         struct mptcp_rm_list alist = { .nr = 0 }, slist = { .nr = 0 };
1207         struct mptcp_pm_addr_entry *entry;
1208
1209         list_for_each_entry(entry, rm_list, list) {
1210                 if (lookup_subflow_by_saddr(&msk->conn_list, &entry->addr) &&
1211                     alist.nr < MPTCP_RM_IDS_MAX &&
1212                     slist.nr < MPTCP_RM_IDS_MAX) {
1213                         alist.ids[alist.nr++] = entry->addr.id;
1214                         slist.ids[slist.nr++] = entry->addr.id;
1215                 } else if (remove_anno_list_by_saddr(msk, &entry->addr) &&
1216                          alist.nr < MPTCP_RM_IDS_MAX) {
1217                         alist.ids[alist.nr++] = entry->addr.id;
1218                 }
1219         }
1220
1221         if (alist.nr) {
1222                 spin_lock_bh(&msk->pm.lock);
1223                 mptcp_pm_remove_addr(msk, &alist);
1224                 spin_unlock_bh(&msk->pm.lock);
1225         }
1226         if (slist.nr)
1227                 mptcp_pm_remove_subflow(msk, &slist);
1228 }
1229
1230 static void mptcp_nl_remove_addrs_list(struct net *net,
1231                                        struct list_head *rm_list)
1232 {
1233         long s_slot = 0, s_num = 0;
1234         struct mptcp_sock *msk;
1235
1236         if (list_empty(rm_list))
1237                 return;
1238
1239         while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) {
1240                 struct sock *sk = (struct sock *)msk;
1241
1242                 lock_sock(sk);
1243                 mptcp_pm_remove_addrs_and_subflows(msk, rm_list);
1244                 release_sock(sk);
1245
1246                 sock_put(sk);
1247                 cond_resched();
1248         }
1249 }
1250
1251 static void __flush_addrs(struct list_head *list)
1252 {
1253         while (!list_empty(list)) {
1254                 struct mptcp_pm_addr_entry *cur;
1255
1256                 cur = list_entry(list->next,
1257                                  struct mptcp_pm_addr_entry, list);
1258                 list_del_rcu(&cur->list);
1259                 mptcp_pm_free_addr_entry(cur);
1260         }
1261 }
1262
1263 static void __reset_counters(struct pm_nl_pernet *pernet)
1264 {
1265         WRITE_ONCE(pernet->add_addr_signal_max, 0);
1266         WRITE_ONCE(pernet->add_addr_accept_max, 0);
1267         WRITE_ONCE(pernet->local_addr_max, 0);
1268         pernet->addrs = 0;
1269 }
1270
1271 static int mptcp_nl_cmd_flush_addrs(struct sk_buff *skb, struct genl_info *info)
1272 {
1273         struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
1274         LIST_HEAD(free_list);
1275
1276         spin_lock_bh(&pernet->lock);
1277         list_splice_init(&pernet->local_addr_list, &free_list);
1278         __reset_counters(pernet);
1279         pernet->next_id = 1;
1280         bitmap_zero(pernet->id_bitmap, MAX_ADDR_ID + 1);
1281         spin_unlock_bh(&pernet->lock);
1282         mptcp_nl_remove_addrs_list(sock_net(skb->sk), &free_list);
1283         __flush_addrs(&free_list);
1284         return 0;
1285 }
1286
1287 static int mptcp_nl_fill_addr(struct sk_buff *skb,
1288                               struct mptcp_pm_addr_entry *entry)
1289 {
1290         struct mptcp_addr_info *addr = &entry->addr;
1291         struct nlattr *attr;
1292
1293         attr = nla_nest_start(skb, MPTCP_PM_ATTR_ADDR);
1294         if (!attr)
1295                 return -EMSGSIZE;
1296
1297         if (nla_put_u16(skb, MPTCP_PM_ADDR_ATTR_FAMILY, addr->family))
1298                 goto nla_put_failure;
1299         if (nla_put_u16(skb, MPTCP_PM_ADDR_ATTR_PORT, ntohs(addr->port)))
1300                 goto nla_put_failure;
1301         if (nla_put_u8(skb, MPTCP_PM_ADDR_ATTR_ID, addr->id))
1302                 goto nla_put_failure;
1303         if (nla_put_u32(skb, MPTCP_PM_ADDR_ATTR_FLAGS, entry->addr.flags))
1304                 goto nla_put_failure;
1305         if (entry->addr.ifindex &&
1306             nla_put_s32(skb, MPTCP_PM_ADDR_ATTR_IF_IDX, entry->addr.ifindex))
1307                 goto nla_put_failure;
1308
1309         if (addr->family == AF_INET &&
1310             nla_put_in_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR4,
1311                             addr->addr.s_addr))
1312                 goto nla_put_failure;
1313 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
1314         else if (addr->family == AF_INET6 &&
1315                  nla_put_in6_addr(skb, MPTCP_PM_ADDR_ATTR_ADDR6, &addr->addr6))
1316                 goto nla_put_failure;
1317 #endif
1318         nla_nest_end(skb, attr);
1319         return 0;
1320
1321 nla_put_failure:
1322         nla_nest_cancel(skb, attr);
1323         return -EMSGSIZE;
1324 }
1325
1326 static int mptcp_nl_cmd_get_addr(struct sk_buff *skb, struct genl_info *info)
1327 {
1328         struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR];
1329         struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
1330         struct mptcp_pm_addr_entry addr, *entry;
1331         struct sk_buff *msg;
1332         void *reply;
1333         int ret;
1334
1335         ret = mptcp_pm_parse_addr(attr, info, false, &addr);
1336         if (ret < 0)
1337                 return ret;
1338
1339         msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
1340         if (!msg)
1341                 return -ENOMEM;
1342
1343         reply = genlmsg_put_reply(msg, info, &mptcp_genl_family, 0,
1344                                   info->genlhdr->cmd);
1345         if (!reply) {
1346                 GENL_SET_ERR_MSG(info, "not enough space in Netlink message");
1347                 ret = -EMSGSIZE;
1348                 goto fail;
1349         }
1350
1351         spin_lock_bh(&pernet->lock);
1352         entry = __lookup_addr_by_id(pernet, addr.addr.id);
1353         if (!entry) {
1354                 GENL_SET_ERR_MSG(info, "address not found");
1355                 ret = -EINVAL;
1356                 goto unlock_fail;
1357         }
1358
1359         ret = mptcp_nl_fill_addr(msg, entry);
1360         if (ret)
1361                 goto unlock_fail;
1362
1363         genlmsg_end(msg, reply);
1364         ret = genlmsg_reply(msg, info);
1365         spin_unlock_bh(&pernet->lock);
1366         return ret;
1367
1368 unlock_fail:
1369         spin_unlock_bh(&pernet->lock);
1370
1371 fail:
1372         nlmsg_free(msg);
1373         return ret;
1374 }
1375
1376 static int mptcp_nl_cmd_dump_addrs(struct sk_buff *msg,
1377                                    struct netlink_callback *cb)
1378 {
1379         struct net *net = sock_net(msg->sk);
1380         struct mptcp_pm_addr_entry *entry;
1381         struct pm_nl_pernet *pernet;
1382         int id = cb->args[0];
1383         void *hdr;
1384         int i;
1385
1386         pernet = net_generic(net, pm_nl_pernet_id);
1387
1388         spin_lock_bh(&pernet->lock);
1389         for (i = id; i < MAX_ADDR_ID + 1; i++) {
1390                 if (test_bit(i, pernet->id_bitmap)) {
1391                         entry = __lookup_addr_by_id(pernet, i);
1392                         if (!entry)
1393                                 break;
1394
1395                         if (entry->addr.id <= id)
1396                                 continue;
1397
1398                         hdr = genlmsg_put(msg, NETLINK_CB(cb->skb).portid,
1399                                           cb->nlh->nlmsg_seq, &mptcp_genl_family,
1400                                           NLM_F_MULTI, MPTCP_PM_CMD_GET_ADDR);
1401                         if (!hdr)
1402                                 break;
1403
1404                         if (mptcp_nl_fill_addr(msg, entry) < 0) {
1405                                 genlmsg_cancel(msg, hdr);
1406                                 break;
1407                         }
1408
1409                         id = entry->addr.id;
1410                         genlmsg_end(msg, hdr);
1411                 }
1412         }
1413         spin_unlock_bh(&pernet->lock);
1414
1415         cb->args[0] = id;
1416         return msg->len;
1417 }
1418
1419 static int parse_limit(struct genl_info *info, int id, unsigned int *limit)
1420 {
1421         struct nlattr *attr = info->attrs[id];
1422
1423         if (!attr)
1424                 return 0;
1425
1426         *limit = nla_get_u32(attr);
1427         if (*limit > MPTCP_PM_ADDR_MAX) {
1428                 GENL_SET_ERR_MSG(info, "limit greater than maximum");
1429                 return -EINVAL;
1430         }
1431         return 0;
1432 }
1433
1434 static int
1435 mptcp_nl_cmd_set_limits(struct sk_buff *skb, struct genl_info *info)
1436 {
1437         struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
1438         unsigned int rcv_addrs, subflows;
1439         int ret;
1440
1441         spin_lock_bh(&pernet->lock);
1442         rcv_addrs = pernet->add_addr_accept_max;
1443         ret = parse_limit(info, MPTCP_PM_ATTR_RCV_ADD_ADDRS, &rcv_addrs);
1444         if (ret)
1445                 goto unlock;
1446
1447         subflows = pernet->subflows_max;
1448         ret = parse_limit(info, MPTCP_PM_ATTR_SUBFLOWS, &subflows);
1449         if (ret)
1450                 goto unlock;
1451
1452         WRITE_ONCE(pernet->add_addr_accept_max, rcv_addrs);
1453         WRITE_ONCE(pernet->subflows_max, subflows);
1454
1455 unlock:
1456         spin_unlock_bh(&pernet->lock);
1457         return ret;
1458 }
1459
1460 static int
1461 mptcp_nl_cmd_get_limits(struct sk_buff *skb, struct genl_info *info)
1462 {
1463         struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
1464         struct sk_buff *msg;
1465         void *reply;
1466
1467         msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
1468         if (!msg)
1469                 return -ENOMEM;
1470
1471         reply = genlmsg_put_reply(msg, info, &mptcp_genl_family, 0,
1472                                   MPTCP_PM_CMD_GET_LIMITS);
1473         if (!reply)
1474                 goto fail;
1475
1476         if (nla_put_u32(msg, MPTCP_PM_ATTR_RCV_ADD_ADDRS,
1477                         READ_ONCE(pernet->add_addr_accept_max)))
1478                 goto fail;
1479
1480         if (nla_put_u32(msg, MPTCP_PM_ATTR_SUBFLOWS,
1481                         READ_ONCE(pernet->subflows_max)))
1482                 goto fail;
1483
1484         genlmsg_end(msg, reply);
1485         return genlmsg_reply(msg, info);
1486
1487 fail:
1488         GENL_SET_ERR_MSG(info, "not enough space in Netlink message");
1489         nlmsg_free(msg);
1490         return -EMSGSIZE;
1491 }
1492
1493 static int mptcp_nl_addr_backup(struct net *net,
1494                                 struct mptcp_addr_info *addr,
1495                                 u8 bkup)
1496 {
1497         long s_slot = 0, s_num = 0;
1498         struct mptcp_sock *msk;
1499         int ret = -EINVAL;
1500
1501         while ((msk = mptcp_token_iter_next(net, &s_slot, &s_num)) != NULL) {
1502                 struct sock *sk = (struct sock *)msk;
1503
1504                 if (list_empty(&msk->conn_list))
1505                         goto next;
1506
1507                 lock_sock(sk);
1508                 spin_lock_bh(&msk->pm.lock);
1509                 ret = mptcp_pm_nl_mp_prio_send_ack(msk, addr, bkup);
1510                 spin_unlock_bh(&msk->pm.lock);
1511                 release_sock(sk);
1512
1513 next:
1514                 sock_put(sk);
1515                 cond_resched();
1516         }
1517
1518         return ret;
1519 }
1520
1521 static int mptcp_nl_cmd_set_flags(struct sk_buff *skb, struct genl_info *info)
1522 {
1523         struct nlattr *attr = info->attrs[MPTCP_PM_ATTR_ADDR];
1524         struct pm_nl_pernet *pernet = genl_info_pm_nl(info);
1525         struct mptcp_pm_addr_entry addr, *entry;
1526         struct net *net = sock_net(skb->sk);
1527         u8 bkup = 0;
1528         int ret;
1529
1530         ret = mptcp_pm_parse_addr(attr, info, true, &addr);
1531         if (ret < 0)
1532                 return ret;
1533
1534         if (addr.addr.flags & MPTCP_PM_ADDR_FLAG_BACKUP)
1535                 bkup = 1;
1536
1537         list_for_each_entry(entry, &pernet->local_addr_list, list) {
1538                 if (addresses_equal(&entry->addr, &addr.addr, true)) {
1539                         ret = mptcp_nl_addr_backup(net, &entry->addr, bkup);
1540                         if (ret)
1541                                 return ret;
1542
1543                         if (bkup)
1544                                 entry->addr.flags |= MPTCP_PM_ADDR_FLAG_BACKUP;
1545                         else
1546                                 entry->addr.flags &= ~MPTCP_PM_ADDR_FLAG_BACKUP;
1547                 }
1548         }
1549
1550         return 0;
1551 }
1552
1553 static void mptcp_nl_mcast_send(struct net *net, struct sk_buff *nlskb, gfp_t gfp)
1554 {
1555         genlmsg_multicast_netns(&mptcp_genl_family, net,
1556                                 nlskb, 0, MPTCP_PM_EV_GRP_OFFSET, gfp);
1557 }
1558
1559 static int mptcp_event_add_subflow(struct sk_buff *skb, const struct sock *ssk)
1560 {
1561         const struct inet_sock *issk = inet_sk(ssk);
1562         const struct mptcp_subflow_context *sf;
1563
1564         if (nla_put_u16(skb, MPTCP_ATTR_FAMILY, ssk->sk_family))
1565                 return -EMSGSIZE;
1566
1567         switch (ssk->sk_family) {
1568         case AF_INET:
1569                 if (nla_put_in_addr(skb, MPTCP_ATTR_SADDR4, issk->inet_saddr))
1570                         return -EMSGSIZE;
1571                 if (nla_put_in_addr(skb, MPTCP_ATTR_DADDR4, issk->inet_daddr))
1572                         return -EMSGSIZE;
1573                 break;
1574 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
1575         case AF_INET6: {
1576                 const struct ipv6_pinfo *np = inet6_sk(ssk);
1577
1578                 if (nla_put_in6_addr(skb, MPTCP_ATTR_SADDR6, &np->saddr))
1579                         return -EMSGSIZE;
1580                 if (nla_put_in6_addr(skb, MPTCP_ATTR_DADDR6, &ssk->sk_v6_daddr))
1581                         return -EMSGSIZE;
1582                 break;
1583         }
1584 #endif
1585         default:
1586                 WARN_ON_ONCE(1);
1587                 return -EMSGSIZE;
1588         }
1589
1590         if (nla_put_be16(skb, MPTCP_ATTR_SPORT, issk->inet_sport))
1591                 return -EMSGSIZE;
1592         if (nla_put_be16(skb, MPTCP_ATTR_DPORT, issk->inet_dport))
1593                 return -EMSGSIZE;
1594
1595         sf = mptcp_subflow_ctx(ssk);
1596         if (WARN_ON_ONCE(!sf))
1597                 return -EINVAL;
1598
1599         if (nla_put_u8(skb, MPTCP_ATTR_LOC_ID, sf->local_id))
1600                 return -EMSGSIZE;
1601
1602         if (nla_put_u8(skb, MPTCP_ATTR_REM_ID, sf->remote_id))
1603                 return -EMSGSIZE;
1604
1605         return 0;
1606 }
1607
1608 static int mptcp_event_put_token_and_ssk(struct sk_buff *skb,
1609                                          const struct mptcp_sock *msk,
1610                                          const struct sock *ssk)
1611 {
1612         const struct sock *sk = (const struct sock *)msk;
1613         const struct mptcp_subflow_context *sf;
1614         u8 sk_err;
1615
1616         if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token))
1617                 return -EMSGSIZE;
1618
1619         if (mptcp_event_add_subflow(skb, ssk))
1620                 return -EMSGSIZE;
1621
1622         sf = mptcp_subflow_ctx(ssk);
1623         if (WARN_ON_ONCE(!sf))
1624                 return -EINVAL;
1625
1626         if (nla_put_u8(skb, MPTCP_ATTR_BACKUP, sf->backup))
1627                 return -EMSGSIZE;
1628
1629         if (ssk->sk_bound_dev_if &&
1630             nla_put_s32(skb, MPTCP_ATTR_IF_IDX, ssk->sk_bound_dev_if))
1631                 return -EMSGSIZE;
1632
1633         sk_err = ssk->sk_err;
1634         if (sk_err && sk->sk_state == TCP_ESTABLISHED &&
1635             nla_put_u8(skb, MPTCP_ATTR_ERROR, sk_err))
1636                 return -EMSGSIZE;
1637
1638         return 0;
1639 }
1640
1641 static int mptcp_event_sub_established(struct sk_buff *skb,
1642                                        const struct mptcp_sock *msk,
1643                                        const struct sock *ssk)
1644 {
1645         return mptcp_event_put_token_and_ssk(skb, msk, ssk);
1646 }
1647
1648 static int mptcp_event_sub_closed(struct sk_buff *skb,
1649                                   const struct mptcp_sock *msk,
1650                                   const struct sock *ssk)
1651 {
1652         if (mptcp_event_put_token_and_ssk(skb, msk, ssk))
1653                 return -EMSGSIZE;
1654
1655         return 0;
1656 }
1657
1658 static int mptcp_event_created(struct sk_buff *skb,
1659                                const struct mptcp_sock *msk,
1660                                const struct sock *ssk)
1661 {
1662         int err = nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token);
1663
1664         if (err)
1665                 return err;
1666
1667         return mptcp_event_add_subflow(skb, ssk);
1668 }
1669
1670 void mptcp_event_addr_removed(const struct mptcp_sock *msk, uint8_t id)
1671 {
1672         struct net *net = sock_net((const struct sock *)msk);
1673         struct nlmsghdr *nlh;
1674         struct sk_buff *skb;
1675
1676         if (!genl_has_listeners(&mptcp_genl_family, net, MPTCP_PM_EV_GRP_OFFSET))
1677                 return;
1678
1679         skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC);
1680         if (!skb)
1681                 return;
1682
1683         nlh = genlmsg_put(skb, 0, 0, &mptcp_genl_family, 0, MPTCP_EVENT_REMOVED);
1684         if (!nlh)
1685                 goto nla_put_failure;
1686
1687         if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token))
1688                 goto nla_put_failure;
1689
1690         if (nla_put_u8(skb, MPTCP_ATTR_REM_ID, id))
1691                 goto nla_put_failure;
1692
1693         genlmsg_end(skb, nlh);
1694         mptcp_nl_mcast_send(net, skb, GFP_ATOMIC);
1695         return;
1696
1697 nla_put_failure:
1698         kfree_skb(skb);
1699 }
1700
1701 void mptcp_event_addr_announced(const struct mptcp_sock *msk,
1702                                 const struct mptcp_addr_info *info)
1703 {
1704         struct net *net = sock_net((const struct sock *)msk);
1705         struct nlmsghdr *nlh;
1706         struct sk_buff *skb;
1707
1708         if (!genl_has_listeners(&mptcp_genl_family, net, MPTCP_PM_EV_GRP_OFFSET))
1709                 return;
1710
1711         skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC);
1712         if (!skb)
1713                 return;
1714
1715         nlh = genlmsg_put(skb, 0, 0, &mptcp_genl_family, 0,
1716                           MPTCP_EVENT_ANNOUNCED);
1717         if (!nlh)
1718                 goto nla_put_failure;
1719
1720         if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token))
1721                 goto nla_put_failure;
1722
1723         if (nla_put_u8(skb, MPTCP_ATTR_REM_ID, info->id))
1724                 goto nla_put_failure;
1725
1726         if (nla_put_be16(skb, MPTCP_ATTR_DPORT, info->port))
1727                 goto nla_put_failure;
1728
1729         switch (info->family) {
1730         case AF_INET:
1731                 if (nla_put_in_addr(skb, MPTCP_ATTR_DADDR4, info->addr.s_addr))
1732                         goto nla_put_failure;
1733                 break;
1734 #if IS_ENABLED(CONFIG_MPTCP_IPV6)
1735         case AF_INET6:
1736                 if (nla_put_in6_addr(skb, MPTCP_ATTR_DADDR6, &info->addr6))
1737                         goto nla_put_failure;
1738                 break;
1739 #endif
1740         default:
1741                 WARN_ON_ONCE(1);
1742                 goto nla_put_failure;
1743         }
1744
1745         genlmsg_end(skb, nlh);
1746         mptcp_nl_mcast_send(net, skb, GFP_ATOMIC);
1747         return;
1748
1749 nla_put_failure:
1750         kfree_skb(skb);
1751 }
1752
1753 void mptcp_event(enum mptcp_event_type type, const struct mptcp_sock *msk,
1754                  const struct sock *ssk, gfp_t gfp)
1755 {
1756         struct net *net = sock_net((const struct sock *)msk);
1757         struct nlmsghdr *nlh;
1758         struct sk_buff *skb;
1759
1760         if (!genl_has_listeners(&mptcp_genl_family, net, MPTCP_PM_EV_GRP_OFFSET))
1761                 return;
1762
1763         skb = nlmsg_new(NLMSG_DEFAULT_SIZE, gfp);
1764         if (!skb)
1765                 return;
1766
1767         nlh = genlmsg_put(skb, 0, 0, &mptcp_genl_family, 0, type);
1768         if (!nlh)
1769                 goto nla_put_failure;
1770
1771         switch (type) {
1772         case MPTCP_EVENT_UNSPEC:
1773                 WARN_ON_ONCE(1);
1774                 break;
1775         case MPTCP_EVENT_CREATED:
1776         case MPTCP_EVENT_ESTABLISHED:
1777                 if (mptcp_event_created(skb, msk, ssk) < 0)
1778                         goto nla_put_failure;
1779                 break;
1780         case MPTCP_EVENT_CLOSED:
1781                 if (nla_put_u32(skb, MPTCP_ATTR_TOKEN, msk->token) < 0)
1782                         goto nla_put_failure;
1783                 break;
1784         case MPTCP_EVENT_ANNOUNCED:
1785         case MPTCP_EVENT_REMOVED:
1786                 /* call mptcp_event_addr_announced()/removed instead */
1787                 WARN_ON_ONCE(1);
1788                 break;
1789         case MPTCP_EVENT_SUB_ESTABLISHED:
1790         case MPTCP_EVENT_SUB_PRIORITY:
1791                 if (mptcp_event_sub_established(skb, msk, ssk) < 0)
1792                         goto nla_put_failure;
1793                 break;
1794         case MPTCP_EVENT_SUB_CLOSED:
1795                 if (mptcp_event_sub_closed(skb, msk, ssk) < 0)
1796                         goto nla_put_failure;
1797                 break;
1798         }
1799
1800         genlmsg_end(skb, nlh);
1801         mptcp_nl_mcast_send(net, skb, gfp);
1802         return;
1803
1804 nla_put_failure:
1805         kfree_skb(skb);
1806 }
1807
1808 static const struct genl_small_ops mptcp_pm_ops[] = {
1809         {
1810                 .cmd    = MPTCP_PM_CMD_ADD_ADDR,
1811                 .doit   = mptcp_nl_cmd_add_addr,
1812                 .flags  = GENL_ADMIN_PERM,
1813         },
1814         {
1815                 .cmd    = MPTCP_PM_CMD_DEL_ADDR,
1816                 .doit   = mptcp_nl_cmd_del_addr,
1817                 .flags  = GENL_ADMIN_PERM,
1818         },
1819         {
1820                 .cmd    = MPTCP_PM_CMD_FLUSH_ADDRS,
1821                 .doit   = mptcp_nl_cmd_flush_addrs,
1822                 .flags  = GENL_ADMIN_PERM,
1823         },
1824         {
1825                 .cmd    = MPTCP_PM_CMD_GET_ADDR,
1826                 .doit   = mptcp_nl_cmd_get_addr,
1827                 .dumpit   = mptcp_nl_cmd_dump_addrs,
1828         },
1829         {
1830                 .cmd    = MPTCP_PM_CMD_SET_LIMITS,
1831                 .doit   = mptcp_nl_cmd_set_limits,
1832                 .flags  = GENL_ADMIN_PERM,
1833         },
1834         {
1835                 .cmd    = MPTCP_PM_CMD_GET_LIMITS,
1836                 .doit   = mptcp_nl_cmd_get_limits,
1837         },
1838         {
1839                 .cmd    = MPTCP_PM_CMD_SET_FLAGS,
1840                 .doit   = mptcp_nl_cmd_set_flags,
1841                 .flags  = GENL_ADMIN_PERM,
1842         },
1843 };
1844
1845 static struct genl_family mptcp_genl_family __ro_after_init = {
1846         .name           = MPTCP_PM_NAME,
1847         .version        = MPTCP_PM_VER,
1848         .maxattr        = MPTCP_PM_ATTR_MAX,
1849         .policy         = mptcp_pm_policy,
1850         .netnsok        = true,
1851         .module         = THIS_MODULE,
1852         .small_ops      = mptcp_pm_ops,
1853         .n_small_ops    = ARRAY_SIZE(mptcp_pm_ops),
1854         .mcgrps         = mptcp_pm_mcgrps,
1855         .n_mcgrps       = ARRAY_SIZE(mptcp_pm_mcgrps),
1856 };
1857
1858 static int __net_init pm_nl_init_net(struct net *net)
1859 {
1860         struct pm_nl_pernet *pernet = net_generic(net, pm_nl_pernet_id);
1861
1862         INIT_LIST_HEAD_RCU(&pernet->local_addr_list);
1863         __reset_counters(pernet);
1864         pernet->next_id = 1;
1865         bitmap_zero(pernet->id_bitmap, MAX_ADDR_ID + 1);
1866         spin_lock_init(&pernet->lock);
1867         return 0;
1868 }
1869
1870 static void __net_exit pm_nl_exit_net(struct list_head *net_list)
1871 {
1872         struct net *net;
1873
1874         list_for_each_entry(net, net_list, exit_list) {
1875                 struct pm_nl_pernet *pernet = net_generic(net, pm_nl_pernet_id);
1876
1877                 /* net is removed from namespace list, can't race with
1878                  * other modifiers
1879                  */
1880                 __flush_addrs(&pernet->local_addr_list);
1881         }
1882 }
1883
1884 static struct pernet_operations mptcp_pm_pernet_ops = {
1885         .init = pm_nl_init_net,
1886         .exit_batch = pm_nl_exit_net,
1887         .id = &pm_nl_pernet_id,
1888         .size = sizeof(struct pm_nl_pernet),
1889 };
1890
1891 void __init mptcp_pm_nl_init(void)
1892 {
1893         if (register_pernet_subsys(&mptcp_pm_pernet_ops) < 0)
1894                 panic("Failed to register MPTCP PM pernet subsystem.\n");
1895
1896         if (genl_register_family(&mptcp_genl_family))
1897                 panic("Failed to register MPTCP PM netlink family\n");
1898 }